The problem with science...

Hi all,

Something has bugged me for a while, and I feel like dumping on you (If you care to listen). This may have been said before so please cough me if so but here is the real problem with science.

It is inaccessible to you.

That's right, you.

Guess what, I found a great article the other day published in PNAS. (This linked article is picked semi randomly to illustrate my point)

http://www.pnas.org/content/108/35/14416.abstract

Try to read the full text. Oh, what sorry, unless you have paid for the subscription/are part of a University Network etc. You can't.

Obviously, you only wanted to read the abstract, and not what was actually done.

Now what really gets me about this is that is that you will almost invariably have to rely on what reporters/ popular science writers tell you about what really happened in the paper. This is like some game of telephone, where the person relaying you information doesn't really speak the same language, and has a hidden motive.

Oh and get this, the researchers have to pay to have the article published in the paper. Fun eh ?!

Ok so here is really the question about America (coming from an outsider, with utter disbelief about your potential presidential candidates).
Would people like this actually exist in a world where knowledge was free?

Imagine for instance:
I don't believe in "XYZ"..
I go educate myself on the subject from data presented by actual scientists (peer reviewed journal articles).
Then I decide.

Do we think this would clarify a lot of these "debates" (evolution/climate change/homosexuality/race)? or am I being naive and would people just be lazy and still rely on what others tell them?

I would suggest that academic articles are among the least convincing forms of communication we have yet devised. They are complicated, rely predominantly on reference to other, equally-complicated documents, usually target a very niche audience, et cetera, et cetera. Probably the more important thing for laymen people (like myself, on the vast majority of subjects) is to expand the circle of trust towards more mainstream sources (that is, being able to read news reports trusting that they are not horrifyingly inaccurate and/or deceptive).

Scientific papers need to get published in peer-reviewed and respected journals. These journals are typically small in circulation and have a niche audience, or are not for profit (like PNAS, which is government sponsored), yet they have to support themselves for both paper and electronic publishing. That costs money, and unless someone is willing to underwrite them, they need to charge the users. And note that PNAS *does* charge authors, because that's part of how they generate operational fees, not because they are greedy bastards.

Universities and companies typically subscribe; you could easily go to a local college, find the latest PNAS on their shelves or on their closed systems, and access the paper there. Or, you could simply wait six months, and the content goes free online. Is that really "inaccessible" or just inconvenient?

Unless you're working in the field, do you really need more than the abstract? And if you do need that, is $10 too much to ask? As noted, the information *will* be free in six months.

Imagine for instance:
I don't believe in "XYZ"..
I go educate myself on the subject from data presented by actual scientists (peer reviewed journal articles).
Then I decide.

Do we think this would clarify a lot of these "debates" (evolution/climate change/homosexuality/race)? or am I being naive and would people just be lazy and still rely on what others tell them?

Unfortunately, we know several things. First, people's minds are not likely to be changed by evidence unless they are predisposed (usually through training) to judge their beliefs by the evidence available and thus consciously change them, even if that disagrees with their feelings, their authority figures, their political beliefs, even their religious beliefs. Secondly, in order to judge the validity of any paper, you need an understanding of it's context in the field; what are current opposing positions, is the logic correct, is the math correct, are there missing elements that would mitigate against it? You even need to understand the reputation of the journal and it's reviewers, and the method of review. I don't think that the majority of Americans are likely trained to do that, or would accept conclusions that are emotionally distasteful to them under any circumstances.

I've found that there *are* some reasonable pop-sci sources that can be used as a starting point for evaluating new research. "Science Daily" is my go-to. But I use the basics of informal logic and searches of other sources, blogs and explanations of established science to judge these releases, if I want to dig into them, and that takes effort as well as an understanding of the methods of science, logic and mathematics. It's not typically *too* difficult, but it's not easy either. And very often, there's a *lot* of context for controversial conclusions, such that it might take you weeks of noodling around and research just to be sure a paper is sufficient to trust, or not.

I don't think you're being naive, I just think that people whose minds can be changed by evidence are predisposed to consider things critically and analytically. Others will trust their friends, their pastor, their boss, their parents, their siblings, even their professors, before they consider something from some person they don't know who disagrees with them. And as you know, there is a *huge* effort in areas of controversy to put fake science into the public eye. (Try "Watt's Up With That" or the hoary "www.junkscience.com" for ongoing examples of deliberately misleading information posing as scientific information.)

Information available online is vastly more easy to get than library information of 20 years ago, yet the debates still continue, and science is *losing* some of them. Unless some foundation is willing to fund free distribution of literally thousands of journals, it's not going to get easier to find this info for free. But at the same time, it's not *that* hard to get, if you want it.

BTW, don't forget Google Scholar. Also, as always, "Sweet Reason - A Field Guide to Modern Logic" by Timoczko, Garfield and Henle (2nd edition imminent) is an awesome intro. (Disclosure - Jay Garfield contributed to both the first and second editions, but is not listed as an author on the first. He was my philosophy professor years ago.)

The whole scientific journal thing is a gigantic fustercluck, to put it mildly. This link on Metafilter is a good starting point for some fascinating, and rather disturbing, reading. The comments have a bunch of other good links as well.

In very short, the journals are making insane profits largely by using volunteer labor. They charge people to publish, they charge people to read, and they get senior scientists to do reviews of papers for free. Their operating margins are about 40%, and a subscription to most journals costs thousands of dollars. Reading a single paper will usually set you back thirty to forty bones.

It is an all-caps RACKET.

Having been in university for 10 years over various degrees i have to say i fall on the side that Malor is on. The costs to peer-review and produce these things is negligible. Why? Because the researchers do all the work - they edit, review, write and perform the experiments. The only thing that the publishers do is store the content (i mean, seriously when was the last time you saw a properly edited paper? There's always spelling mistakes - especially from people whose primary language is not English and this isn't in small papers, this is for Elselvier and co.) and they also keep contact with other academics so that they can provide the reviewing process.

While i'm all for charging for the content it should be a couple of $/£/€ and not £15-30 per article - any of which can range from 2-3 pages to a whopping 10-20 and yet they're the same price.

I think there are two separate discussions here.

1. Would greater availability of scientific data clarify a lot of these "debates" (evolution/climate change/homosexuality/race)? There's plenty of indication that many people are just not interested in an evidence-based discussion of these topics and seek only to advance a pre-determined agenda, based on a beliefe system, regardless of data. Data is, and will continue to be, heavily cherry-picked. Far too much religion in the way for a true public discussion of those topics.

2. The accessibility of research, publicly-funded and supposedly for the good of all.

The biggest thing I've learned since I started working in healthcare is that "scientific" white papers are nothing more than glorified marketing brochures. The data is so massaged and filtered to meet the aims of the white paper. For every clinical implant I can show you 5 white papers that say the product is the best that's ever been invented and 5 that say it's garbage.

The "science" is VERY subjective.

Bear wrote:

The biggest thing I've learned since I started working in healthcare is that "scientific" white papers are nothing more than glorified marketing brochures. The data is so massaged and filtered to meet the aims of the white paper. For every clinical implant I can show you 5 white papers that say the product is the best that's ever been invented and 5 that say it's garbage.

The "science" is VERY subjective.

So would you say that this opinion is related only to Medical Science or Science in general?

PAR

My buddy at NIST always tells me to "beware the layman's explanation". He's not trying to be an elitist, but it often ends up coming across that way. His point is a good one though. It is that some concepts just can't be properly understood without a background in a body of knowledge that can't be obtained with a part time hobbyist's effort and commitment.

My wife tells me pretty much the same thing every time someone comes into her clinic after having watched a pharma advertisement on television and asks for a particular drug. She's convinced "fibromyalgia" is a condition invented by drug companies. Doctors call this the WebMD Phenomenon and say that just about every first year med student ends up with acute cases of WebMD Hypochondria.

Scientific papers are written to present a theory to other scientists, not to convince the layman of anything. Ideally, the journalists who report on science can bridge that gap. Ideally. If they do this for a living and still screw it up so much, how is the layman supposed to understand scientific studies on their own?

It's also the case that scientists are often very smart people who have trained their entire lives to understand their field and those papers. Some of them understand things that I could never quite process as a layman myself (I'm looking at you, mathematicians). We also lack the skills to properly evaluate a scientific theory and judge whether it is more or less sound than another, other than how things feel in our "gut" or how they appeal to "common sense."

Finally, if it was a free-for-all, that's what you'd get. Respected scientific journals cluttered with woo, logical fallacies, and unsupported, wild guesses. Researched papers from the greatest minds in the world, right alongside ranted theories based on common sense, which is defined as something that the author believes to be true but couldn't honestly explain to you why.

However this does NOT mean that science is exactly equivalent to religion, in that people who believe in science are just taking on faith what the priest in the white lab coat hands down to us from the science gods. We know there are checks and balances in place. We know that if the science priest does something stupid or wrong, another science priest proves that. We also know that scientists rarely have things 100% correct, but they're as close as we can get right now, and when it becomes possible to get closer, we'll move closer. It's better to be a little wrong 1,000 times and right once than totally wrong once.

The problem is not with science. The problem is with us.

par wrote:
Bear wrote:

The biggest thing I've learned since I started working in healthcare is that "scientific" white papers are nothing more than glorified marketing brochures. The data is so massaged and filtered to meet the aims of the white paper. For every clinical implant I can show you 5 white papers that say the product is the best that's ever been invented and 5 that say it's garbage.

The "science" is VERY subjective.

So would you say that this opinion is related only to Medical Science or Science in general?

PAR

It's hard for me to say, all of the white papers I've seen have pertained to medical products or devices. What I have seen is that the "outcome" data varies greatly buy the product and company who sponsored the study. In short, if it's your company and product, you have a white paper that shows how AMAZING it is. Meanwhile, your competitor has multiple white papers that say their product is "clinically superior" and yours is probably going to kill people.

Yup.

On a personal note, I helped my wife with her PhD dissertation. She's not a native speaker and needed help writing and editing it. I was there from the beginning so there was a great deal I learned through the process. That said, there were numerous times when I would correct for grammar and/or usage and she would come back and tell me I had completely missed the point of sentence or paragraph and had no idea what I was talking about.

Before I got married, I used to think I was kind of smart. Now I'm just the guy who lifts the heavy stuff.

LobsterMobster wrote:

The problem is not with science. The problem is with us.

That good sir, IS the crux of the problem.

As an aside, any papers that are funded solely, or in part by the NSF or NIH are required to be made publicly available for free..

I have to echo the thought that whilst i think people should be able to access these papers (i'd love to be able to afford a subscription at several journals) and the flow of information will make the world a better place for more informed debate.... The problem is that 90% of the people who read something like that won't understand it at all.

I remember one of my colleagues, not a stupid guy, had read about this new telescope in space "that could see through stars" to be able to see farther back into the universe's history. I told him on the spot i was a bit dubious about those claims but couldn't specifically debunk it because i didn't have the internet to hand.

IIRC, i checked at home and the telescope in question was an telescope designed to use gravitational lensing to see further into space (or something along those lines).

These concepts are just hard to understand from a cursory glance. It's like anything in life - you can either dip into everything and know a little or you can specialise and become an expert. You have to trust the experts in order to continue on living whether they be agriculturalists, stock brokers, engineers etc. For some reason science in general seems to lack this trust from the general populace... perhaps specifically because so much of it is intangible.

Duoae wrote:

I have to echo the thought that whilst i think people should be able to access these papers (i'd love to be able to afford a subscription at several journals) and the flow of information will make the world a better place for more informed debate.... The problem is that 90% of the people who read something like that won't understand it at all.

I remember one of my colleagues, not a stupid guy, had read about this new telescope in space "that could see through stars" to be able to see farther back into the universe's history. I told him on the spot i was a bit dubious about those claims but couldn't specifically debunk it because i didn't have the internet to hand.

IIRC, i checked at home and the telescope in question was an telescope designed to use gravitational lensing to see further into space (or something along those lines).

These concepts are just hard to understand from a cursory glance. It's like anything in life - you can either dip into everything and know a little or you can specialise and become an expert. You have to trust the experts in order to continue on living whether they be agriculturalists, stock brokers, engineers etc. For some reason science in general seems to lack this trust from the general populace... perhaps specifically because so much of it is intangible.

I think part of that mistrust is due to the fact that science is pervasive and offers special insight into just about everything. As a result, its job is, in part, to reveal the falsity or veracity of "time honored truths" in other fields. No one likes when their ox is gored. In particular theologians (whose "field" rests on the shakiest of shaky ground anyway).

I was even thinking of the love-hate relationship people have with the medical profession....

Personally I just think it would be beneficial to all humankind if evidence was available free of charge.

I mean, I consider myself extremely fortunate (have been/still am) a part of a University Network, and therefore have access to the majority of peer-reviewed journals.

It just really bugs me that without this, there is this huge gap in what I can actually find out about reality.

Reading a single paper will usually set you back thirty to forty bones.

Publishing can cost a lot. A review paper my former group put out cost $800 to publish, In a rather low end journal. So yes, I do think it is pretty messed.

I wouldn't mind having to pay the journal in order to publish, if it was open access. But, paying on both ends just feels wrong.

Academic publishing is a racket in general, but I don't know that making this stuff available to the public would change anything. How many people who have an opinion on politics watch C-SPAN? Time is limited, and the papers you're talking about are generally way over the layman's head.

Outside of a library visit to dig up a journal, some fields have membership organizations that grant access to a good portion of relevant published research. For Computer Science, between the IEEE and ACM you'll have access to pretty much every paper you'd care about, and their online archives go back to the 60s or earlier. This isn't true of every field and the memberships aren't free, but it's a far cry from dropping $30-120 on a single research paper that I've heard of elsewhere.

I mostly deal with biology, so maybe this is why I would like the see the field open up the the public more. There are a whole lot of whack jobs attacking the science, that perhaps opening it up would encourage more people to read the actual evidence.

I definitely don't think that it would change everyones minds but if even a small fraction of people who were really interested in the subject had access to it, well I think that would be awesome.
However, we would still get people misrepresenting scientists views (which we still do anyways).

Tanglebones wrote:

As an aside, any papers that are funded solely, or in part by the NSF or NIH are required to be made publicly available for free..

SOCIALISM! IMAGE(http://www.gamerswithjobs.com/files/smileys/icon_oldman.gif)

There's a certain amount of inertia here, I think. Think back to the 50s, and imagine printing documents that include difficult to typeset mathematics, graphs, etc. You're printing them and distributing them to a very limited audience. In that world, of course publication is hard and expensive, and the cost per copy is high as it is not offset by a mass market. As time has progressed, though, things have become less difficult. First, the use of new printing technologies makes typesetting mathematics significantly easier. And now, with the web, the only thing that ought to be an issue any more is the organization required to review papers for publication. (You do still need trusted organizations and trusted individuals to do that sort of thing, to distinguish peer-reviewed work from other work.)

It's a slow boat to move. And you also have requirements on individuals to publish in peer-reviewed journals and present at conferences (which often offset costs partially on the sales of proceedings.)

I think computer science is in pretty good shape right now--at least in PL theory, most of the important works I've sought are available freely online. I suspect the very closeness CS has with technology makes online publication of both draft work and published work rather a no-brainer. Also, the PL theory community's output is pretty small compared to other disciplines, particularly those outside of CS. And things could definitely be better--in particular, the journals make it much easier to keep track of new stuff rather than having to hunt it down. Most people are not inclined to delve as deep as necessary to figure out which papers are important, much less read them all.

It takes a lot of background to understand how a paper on the impact of cosmic rays on climate fits into the big picture, for example. And even if a paper explicitly details the wider implications of the results, that's just fodder for journalists to make a hash of things by misunderstanding. (Hint: "This is a tremendous breakthrough!" means something different than you might expect when you're talking to somebody working on a narrow specialization of a sub-field of a discipline. :D)

complexmath wrote:

Academic publishing is a racket in general, but I don't know that making this stuff available to the public would change anything. How many people who have an opinion on politics watch C-SPAN? Time is limited, and the papers you're talking about are generally way over the layman's head.

Outside of a library visit to dig up a journal, some fields have membership organizations that grant access to a good portion of relevant published research. For Computer Science, between the IEEE and ACM you'll have access to pretty much every paper you'd care about, and their online archives go back to the 60s or earlier. This isn't true of every field and the memberships aren't free, but it's a far cry from dropping $30-120 on a single research paper that I've heard of elsewhere.

That's a good point, as an Aerospace Engineer my membership with the AIAA also gets me access to a large amount of material for much cheaper than if I bought each paper separately, or even subscribed to each publication separately.

And even if a paper explicitly details the wider implications of the results, that's just fodder for journalists to make a hash of things by misunderstanding.

Especially when there are a number of high profile sources you can consult as a journalist who are invested in "misunderstanding" this stuff. Even some journalists have taken sides; The Register does science pieces, but their guy who does climate stories uses Steve Milloy as a primary source and is unabashedly biased.

HantaXP wrote:

Obviously, you only wanted to read the abstract, and not what was actually done.

To be frank, that is often all you need to read

HantaXP wrote:

Oh and get this, the researchers have to pay to have the article published in the paper. Fun eh ?!

Most groups I've worked with have had policies against publishing in journals that ask for up front costs.

HantaXP wrote:

Now what really gets me about this is that is that you will almost invariably have to rely on what reporters/ popular science writers tell you about what really happened in the paper. This is like some game of telephone, where the person relaying you information doesn't really speak the same language, and has a hidden motive.

...

Ok so here is really the question about America (coming from an outsider, with utter disbelief about your potential presidential candidates).
Would people like this actually exist in a world where knowledge was free?

Imagine for instance:
I don't believe in "XYZ"..
I go educate myself on the subject from data presented by actual scientists (peer reviewed journal articles).
Then I decide.

Well this isn't a viable route to understanding a topic for most interested/invested people. Most lay people won't have the skills to understand and interpret a great deal of published science. Hell, most researchers outside of a given micro-discipline often won't have the skills to understand papers published on said topic. But science journalism is not the only route to getting good summaries of the state-of-the-art in much research. Almost all branches of research are in the habit of producing review articles which will summarise the current understanding in a topic and these are usually higher level, written for broader appeal and quite approachable/readable for non-experts. Alongside this, projects like the Cochrane Collaboration, which specialise in meta-analyses of public health research, publish collated re-analyses of data which which, with fairly modest understanding of stats, many people would find approachable.

HantaXP wrote:

Do we think this would clarify a lot of these "debates" (evolution/climate change/homosexuality/race)? or am I being naive and would people just be lazy and still rely on what others tell them?

Access to original research for all won't tangibly help when most people don't have the skills to understand the research. I do think you are being naive that people, even those with the skills, would all rush to the primary research to understand very complex, high-level phenomena. Open access to primary research is important not so much because everybody will actually go back and check facts but because it allows, those (few) with the skills & interest, to go back and fact check review articles, meta-analyses and science journalism pieces.

In my experience as a chemist and materials scientist, papers are more often than not unavailable for the public to scrutinize at any reasonable price point. With regards to:

DanB wrote:

Open access to primary research is important not so much because everybody will actually go back and check facts but because it allows, those (few) with the skills & interest, to go back and fact check review articles, meta-analyses and science journalism pieces.

This is what i believe open or cheap access will allow. It will also allow better information to the public in general... mostly because mainstream reporting of science is absolute rubbish.

Thing is, even more mainstream publications like "Science" or "Nature" (neither of which i particularly rate) are pretty expensive for a subscription.

WRT to the OP; I'm pretty sure PNAS articles go open access 6 months after publication (and it's been that way since 2004). The great majority of UK based science publishers are already open access. It's an increasingly common requirement of government research funding that publication is in open access journals (public taxes paid for the research therefore the public should have access to it). All the major journals in my field are open access. In the vast majority of (bioscience) fields you could be current enough without paying for journal access. And that is to say nothing of the fact that I've yet to meet a paper author who wouldn't send you the pdf if you asked them directly for a copy.

There certainly are issues with academic publishing around the profit margins and free labour but public access is simply not one of the issues today.

Duoae wrote:

In my experience as a chemist and materials scientist, papers are more often than not unavailable for the public to scrutinize at any reasonable price point.

Not even "open access after X months"?

Duoae wrote:

Thing is, even more mainstream publications like "Science" or "Nature" (neither of which i particularly rate) are pretty expensive for a subscription.

I don't think Science and Nature count as mainstream; they certainly aren't lay-public level.

DanB wrote:

It's not just CS a great deal (probably the vast majority) of bioscience research is freely available online too.

Good to know. I didn't want to speak to other disciplines, since I have very little experience with research publications outside of CS.

DanB wrote:
Duoae wrote:

In my experience as a chemist and materials scientist, papers are more often than not unavailable for the public to scrutinize at any reasonable price point.

Not even "open access after X months"?

Do you use anything from Sciencedirect or Wiley? JACS is also limited... I'm interested in the journals that you're accessing for free. Nothing in my field is freely available and that covers a fairly wide variety of stuff.

I wonder if there's IP filters on these sites too? Obviously, whilst using my uni's proxy and within the uni itself i had "free" access to almost everything under the sun.... But once outside and no longer able to use the proxy I was unable to get access to anything I had previously enjoyed. Now i'm in a different country and i wonder if they restrict country-based IPs too...

DanB wrote:
Duoae wrote:

Thing is, even more mainstream publications like "Science" or "Nature" (neither of which i particularly rate) are pretty expensive for a subscription.

I don't think Science and Nature count as mainstream; they certainly aren't lay-public level.

Compared to "proper" articles and journals they are pretty mainstream

[edit]

DanB wrote:

This!
http://www.doaj.org/

No offense to the Polish or Slovenian amongst us.... But i don't class getting access to journals based in those countries (probably a very low exposure index) with something like the Royal Society of Chemistry, Journal of the American Chemical Society, Journal of Alloys and Compounds or Acta Materialia...